This application proposes a five year continuation of an ongoing longitudinal study of attachment working models in adulthood. The current project has yielded significant empirical results and conceptual advances. The proposed project will build on these results to clarify key issues regarding the structure of modern attachment theory: (1) the stability of attachment working models in relation to marriage, collaboration in parenthood, and stressful life events, (2) the relation of working models of subjects' relationships with their parents to working models of their marital relationship at different points in marriage. (3) the relation of working models to marital behavior, partner behavior and partner working models, and differences in how working models relate to marital behavior at different phases of marriage, (4) relations between attachment-theoretical and cognitive-behavioral perspectives on marriage. (5) mechanisms underlying congruence of maternal attachment security and infant attachment security, (6) relations of infant attachment and parenting practices to both maternal and paternal working models, and (7) the cognitive structure of attachment representations. The subjects are 136 married couples recruited as engaged couples 3 months prior to marriage (Phase I) and seen 18 (Phase II) and 36 months (Phase III) later. In the first year of the proposed grant period we will complete the last of our 36 month followups and continue assessments of infant-mother and infant-father attachment. We will conduct home visits to examine secure base behavior with each parent, and also maternal sensitivity. We will also begin to test experimentally the hypothesis that attachment working models are represented in the form of domain specific, temporal-causal narrative sequences (scripts of secure base interactions). Beginning in the second year of the grant period we will begin conducting follow-up attachment (3-year old children) and couples assessments (5 years after Phase I). Assessments of children and couples would continue throughout the funding period. These follow-ups are important because, as explained below, attachment-marriage relations may be differently organized at different stages in marriage and such relations may not be evident in concurrent and short-term follow-ups. They also allow us to examine the magnitude and content of relations between working models and marital behavior into the early years of parenting, and to explore relations of adult and child attachment status with child secure base behavior, maternal sensitivity, marital status, and cooperation with parenting partnership. In brief, the proposed project will provide significant clarifications of the working model concept and the notion that cognitive representations of the child-parent relationship influence later attachment related cognitions and behavior. It -will also help bring interest in working models and secure base behavior into the mainstream of marital theory, and research, where implications for therapy are more likely to flourish.